Metal Reviews

Newest and Best Metal Reviews!
FAQ :: Search :: Members :: Groups :: Register
Login
It is currently Fri Jul 04, 2025 12:28 pm



Post new topic Reply to topic  [ 13 posts ] 
Author Message
 Post subject: Mecha-morals: the ethics of artificial intelligence
PostPosted: Fri Oct 26, 2007 5:25 pm 
Offline
Metal Slave
User avatar

Joined: Thu Sep 16, 2004 3:36 am
Posts: 83
Location: Vancouver, Canada
Here's my second feature, published today. I hope this one stirs a bit of debate!

http://www.ubyssey.bc.ca/2007/10/26/mecha-morals/


Top
 Profile  
 
 Post subject:
PostPosted: Fri Oct 26, 2007 6:22 pm 
I kinda think it's silly for various reasons.

1: It will never be possible to create an AI equivalent to the human brain, or simulate human behaviour for that matter. Not only is it impossible to program as such for various reasons, but there will never be such a powerful computer either.

2: there are various forms of AI. AI doesn't necessarily mean they have (real) emotions. Why would anybody create a robot with psychical and physical emotions anyway?! Sure, you could have robots say "oh, i love you". I can program my Texas instrument calculator to do that too, yet still.. i would have no trouble throwing in the garbage can one day. You can still program an AI to be creative (i highly daubt it will be anywhere near as creative as a human brain though), and "caring" or "sexy" if that's what needed.
AI will always be a discount version of a real person. Had they real emotions they would commit suicide, rape and burn churches like real human beings. So natural you would have to limit them to be nothing more than smart machines with beneficial functions.


Sure i can see the moral issue from an atheistic point of view. I don't believe there is such thing as a soul, so in that sense that makes us kinda equal to higly sophisticated AI. But, in the end of the day an AI will never be anything but 0's and 1's thrown about in different combinations

unless, of course, scientist one day discover a method on how to create syntethic brain tissue for each an every part of the human brain, but i highly daubt that would ever happens.


robots are tools nothing more... Hammers don't have any speciel rights either.

besides... if you could program a highly sophisticated human like AI in a video game... would it be wrong of you to shut down the computer then, or simply quit the game for that matter?


Top
  
 
 Post subject:
PostPosted: Sat Oct 27, 2007 12:14 am 
Offline
Metal King
User avatar

Joined: Tue Jun 06, 2006 11:25 am
Posts: 928
Location: Serres [Greece]
It's a nice article. Maybe because I was expecting more, I was a bit dissappointed to find out that it only scratches the surface of the subject and does not go very deep or into details. It raises some interesting questions though which could develop into larger and more expanded and encompassing essays/articles.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Oct 27, 2007 12:21 am 
Offline
Metal King
User avatar

Joined: Tue Jun 06, 2006 11:25 am
Posts: 928
Location: Serres [Greece]
Also, Astaroth, I think you were very harsh in your judgement and careless in your assumptions. They way I understand them, most (if not all) of your arguments above are flawed or simply lacking, each for different reasons. I would like to argue about them but unfortunately I'm very short of time during these days so I might do that later in the following week.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Oct 27, 2007 1:16 am 
Offline
Metal King

Joined: Sun Aug 19, 2007 2:30 am
Posts: 1212
Antonakis wrote:
Also, Astaroth, I think you were very harsh in your judgement and careless in your assumptions. They way I understand them, most (if not all) of your arguments above are flawed or simply lacking, each for different reasons. I would like to argue about them but unfortunately I'm very short of time during these days so I might do that later in the following week.

This is a nice reply. Maybe because I was expecting more, I was a bit dissappointed to find out that it only scratches the surface of the subject and does not go very deep or into details. It raises some interesting questions though which could develop into larger and more expanded and encompassing replies/flamewars.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Oct 27, 2007 1:52 am 
he he
I'm sorry, Antonakis, i'm not into sci-fi. But i will to try speak in a lighter tone about robots the next time :) wouldn't want to hurt their feelings

but you're welcome to argue with me, and come with your arguments.


anywho, it's still silly... i basicly a question about human sentimentality and our effection for creatures that shows resembles to ourselves. I bet you would threat a sophisticated AI inside a human like robot different from one inside a toaster without a face, or inside a computer game. It's the same fucking thing. AI means artificial intelligence, hence it is nothing but a simulation, be it a NPC in a video game or an AI hooked up on a machine with hydraulics. It's not real and should not have any speciel rights. Software is fucking software.

" But... but... it is not right for a machine to work 24 hours a day, it has the right to freedom, money and 15 hours of rest a day and two days off a week :sad: .. it is not right to use robots as slaves.. bohoo... that it also why i only use my car 7 hours a day"




yeah.. a bit harsh perhaps :lol: but it suits my style


Top
  
 
 Post subject: Re: Mecha-morals: the ethics of artificial intelligence
PostPosted: Sat Oct 27, 2007 9:39 am 
Offline
Metal King
User avatar

Joined: Wed Oct 19, 2005 7:05 pm
Posts: 1421
Location: Brighton
Jaden wrote:
Here's my second feature, published today. I hope this one stirs a bit of debate!

http://www.ubyssey.bc.ca/2007/10/26/mecha-morals/

Thats interesting stuff. nice artical. Think you might find this interesting, relates to your article a bit. You should check out south korean AI if you've not already, some of its breath taking.

http://news.bbc.co.uk/1/hi/technology/6425927.stm

having robot laws like in avimov books would freak me out lol.

Astaroth wrote:
It will never be possible to create an AI equivalent to the human brain, or simulate human behaviour for that matter. Not only is it impossible to program as such for various reasons, but there will never be such a powerful computer either

i'm not sure about that. When you think that we only use around 10% of our brains you don't necessarily need to create an AI of equivalence to the human brain to create something of equal inteligence. Also, the dream of many Scientists is to create software that is truly artificially intelligent, the idea i suppose is that it would be able to create its own programming.

The science behind computers powerfull enough to power such a creation is well underway to. Ever heard of quantum computers?

http://computer.howstuffworks.com/quantum-computer.htm

Give it 10 years and i bet they'll reinvent what possible with computers. quantum computers have the potential to be as powerfull, if not more so then a human brain, its just a matter of how long it takes to get to that stage.


Top
 Profile  
 
 Post subject: Re: Mecha-morals: the ethics of artificial intelligence
PostPosted: Sat Oct 27, 2007 10:39 am 
stuartn15ted wrote:
Astaroth wrote:
It will never be possible to create an AI equivalent to the human brain, or simulate human behaviour for that matter. Not only is it impossible to program as such for various reasons, but there will never be such a powerful computer either

i'm not sure about that. When you think that we only use around 10% of our brains you don't necessarily need to create an AI of equivalence to the human brain to create something of equal inteligence. Also, the dream of many Scientists is to create software that is truly artificially intelligent, the idea i suppose is that it would be able to create its own programming.

The science behind computers powerfull enough to power such a creation is well underway to. Ever heard of quantum computers?

http://computer.howstuffworks.com/quantum-computer.htm

Give it 10 years and i bet they'll reinvent what possible with computers. quantum computers have the potential to be as powerfull, if not more so then a human brain, its just a matter of how long it takes to get to that stage.


interesting... but the deal about using only 10% of the brain is not true. It is an old "myth". Of course, it will also depend on the task that is given. Solving algebra in the head while playing football and listen to music would activate more braincells than just lying on the couch.

I don't believe the 10 years span, neither do they in your article. But it may be possible who knows. But it certainly won't be with our modern computer. Modern computers are only capable of making one calculation at a time (unless they have multiple cores?!). The brain has about 100 billion neurons. Each neuron is connected to 7000 other neurons on average via synapses. They can can all speak to each other simultaneously, not only 1-4 calculations at a time like our modern computer and each neuron can be activated 200 times per second. Even with only 10% activation of the brain it would be impossible to simulate in 10 years.
I know scientist have been able to simulate half a mouse brain on a super computer. It had half of the 8 millions neurons a real mouse have and about 6300 synapses per neuron. The simulation was run for 10 seconds which was equivalent to 1 sec in real life. However, the neuron structure was not interily correct the scientist said.
And personally i can't see how one are capable of simulate a mouse brain. Wouldn't that require a mouse psychiatrist or an understanding of the true nature of the reptilian brain?!

Concerning the human AI, there's the same problem. If you want to create a real human like AI you need to know the true nature of man, psychological. Furthermore, you need to know the different roles each section of the brain has. The brain is not even fully mapped, there still alot to be understood and researched
This is only important if you want to create a realistic humanized AI, of course. And i personally can't see any use of that. It would make more sense create an AI without so-called feelings unless you are a loser who can't get a date. Besides, feelings both psychological and physiological are only meant to keep us alive and make us reproduce. A so-called robot only needs basic instincts to be kept alive or instructions if you like, so i won't go play on the railroad.
And if you want to have a smart AI that can help out humanity with its superior creativity in the production department, why not just keep the AI on the computer?!... Sophisticated computer programs are already helping us out technologically, there's no reason to create smart robot for that.


Top
  
 
 Post subject: Re: Mecha-morals: the ethics of artificial intelligence
PostPosted: Sat Oct 27, 2007 12:50 pm 
Offline
Metal King
User avatar

Joined: Wed Oct 19, 2005 7:05 pm
Posts: 1421
Location: Brighton
Astaroth wrote:
interesting... but the deal about using only 10% of the brain is not true. It is an old "myth". Of course, it will also depend on the task that is given. Solving algebra in the head while playing football and listen to music would activate more braincells than just lying on the couch.

lol, learning all the time.

Astaroth wrote:
I don't believe the 10 years span, neither do they in your article. But it may be possible who knows. But it certainly won't be with our modern computer. Modern computers are only capable of making one calculation at a time (unless they have multiple cores?!). The brain has about 100 billion neurons. Each neuron is connected to 7000 other neurons on average via synapses. They can can all speak to each other simultaneously, not only 1-4 calculations at a time like our modern computer and each neuron can be activated 200 times per second. Even with only 10% activation of the brain it would be impossible to simulate in 10 years.

Christ didn't mean they would be able to recreate the human brain in 10 years. I meant that i think in 10 years time they will have solved many of the problems they're facing now, and the technology will start to change the way computers are used. Its moving along so fast who knows what will be possible by 2017, but i'm pretty such they won't be simulating human brains :lol:

Astaroth wrote:
I know scientist have been able to simulate half a mouse brain on a super computer. It had half of the 8 millions neurons a real mouse have and about 6300 synapses per neuron. The simulation was run for 10 seconds which was equivalent to 1 sec in real life. However, the neuron structure was not interily correct the scientist said.
And personally i can't see how one are capable of simulate a mouse brain. Wouldn't that require a mouse psychiatrist or an understanding of the true nature of the reptilian brain?!

Concerning the human AI, there's the same problem. If you want to create a real human like AI you need to know the true nature of man, psychological. Furthermore, you need to know the different roles each section of the brain has. The brain is not even fully mapped, there still alot to be understood and researched
This is only important if you want to create a realistic humanized AI, of course. And i personally can't see any use of that. It would make more sense create an AI without so-called feelings unless you are a loser who can't get a date. Besides, feelings both psychological and physiological are only meant to keep us alive and make us reproduce. A so-called robot only needs basic instincts to be kept alive or instructions if you like, so i won't go play on the railroad.
And if you want to have a smart AI that can help out humanity with its superior creativity in the production department, why not just keep the AI on the computer?!... Sophisticated computer programs are already helping us out technologically, there's no reason to create smart robot for that.

I don't rly know why anyone would want to make human robots or whatever. Would be pretty cool tho. Be a lot of effort to go to for something that in the end would be pretty gay.

Theres been millions of sci fi books about weather or not it would be possible to create robots that can think like humans or even think they are humans. I have no idea weather it will ever be done but i think that it is possible, just not in our life times.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Oct 27, 2007 3:03 pm 
don't disagree with you...


anywho... wouldn't it be easier to genetically create humans beings without any self-aware ness or self reflection? :P organic robots of some kind?!...

According to Howard Garthner there's multiple intelligences. Here they are, and how i see them fit into my organic "robot":

Bodily-Kinesthetic - should be minimized. It would make a rather retarded robot
Interpersonal - only medium. It is the ability to speak with other ppl and empathy. Wouldn't want them to be too humanistic, else they might get funny ideas
Linguistic - should be somewhat high. But too not too high
Logical-Mathematical - Should be high as fuck
Naturalistic - bah
Spatial - Should also be somewhat high as fuck
Musical - Also high
Intrapersonal - this intelligence should be non-existing. By doing so you could have pretty intelligent and creative ppl you could treat like crap.. .cool!!! Without self-awareness they wouldn't start doing silly stuff like pleading for their life or think about their own existence

furthermore, we should strip them from any primal instincts. Furthmore, we should de-activate their facial expression, so we wouldn't go all sentimental on them. For further improvement of my non-relationship mechanism between real ppl and my organic robots you could also make them african and print a swastika in their forehead.
Now... this is much more easy than producing a computer robot. And if you like to have a strong "robot", you could cut off their arm and re-attach fancy machinery. This is how we should make up for the decreasing population and it's need for ppl to work, or creativity

I believe i'm worthy of the "Albert Einstein World Award of Science" award for my idea... either that or i'm going to steal Al Gore stupid Nobel award... bah...


Top
  
 
 Post subject:
PostPosted: Sun Oct 28, 2007 12:34 am 
Offline
MetalReviews Staff
User avatar

Joined: Wed Feb 09, 2005 4:02 pm
Posts: 29896
Location: UK
I think there'll be a line that gets crossed, before which robots are seen as robots and after which they're seen as being equal. I don't think the second stage will happen in any of our lifetimes though, like Kim said, human clones are a much more pressing issue.

Good article, though, Trevor!


Top
 Profile  
 
 Post subject:
PostPosted: Sun Oct 28, 2007 2:25 am 
Offline
Metal Slave
User avatar

Joined: Thu Sep 16, 2004 3:36 am
Posts: 83
Location: Vancouver, Canada
Astaroth wrote:
I kinda think it's silly for various reasons.

1: It will never be possible to create an AI equivalent to the human brain, or simulate human behaviour for that matter. Not only is it impossible to program as such for various reasons, but there will never be such a powerful computer either.


You're wrong on both accounts.

First, you don't really give any of these supposed "reasons" why such a thing would never be possible. Humans work on processors too, so I fail to see why there is a limitation on this. I would say more, but you never gave any particular reasons for your statement.

And for your second point, again, I don't know why you would even think such a thing. Especially with the way processor speeds are increasing so rapidly and expontentially.

Humans use massive paralell processing (in the trillions), while machines use one - maybe two in some cases. However, Individual computer processors are way faster than individual human processors. It compensates to some degree. Currently humans are quite a bit more powerful than computers, but there's no reason to think this will always be the case. If there's one technology that isn't seeing a stand-still right now, it's increase in computation speeds.

If I recall, the processing power of a computer is comparable to that of a mouse (the accumulated sum of its paralell processing). This may not sound impressive, but it is. Consider the evolutionary timescale. Consider the time between the common ancestor of all living beings and the first creature that was as complex as a mouse. It's massive! So, we've come a long way. I mean, our common ancestor would have also been more complex than our first machine too, so you can even expand the analogy further back a bit. As far as processing power is concerned, I would say we're over half way there. Even if we can't replicate trillions of paralell processes any time soon, we can make up for that with faster processors.



Quote:
2: there are various forms of AI. AI doesn't necessarily mean they have (real) emotions.


Why would one emotion be real and another fake? What does it matter if the entity is made of metal or meat? Emotion is an object of intelligence, which is an object of processes.

Quote:
Why would anybody create a robot with psychical and physical emotions anyway?! Sure, you could have robots say "oh, i love you". I can program my Texas instrument calculator to do that too, yet still.. i would have no trouble throwing in the garbage can one day. You can still program an AI to be creative (i highly daubt it will be anywhere near as creative as a human brain though), and "caring" or "sexy" if that's what needed.


This is just an expansion of what you just said. Again, you don't prove to me that your discrimination is anything but entirely arbitrary.


Quote:
AI will always be a discount version of a real person. Had they real emotions they would commit suicide, rape and burn churches like real human beings. So natural you would have to limit them to be nothing more than smart machines with beneficial functions


If human beings are so bad, then maybe we should stop recreating ourselves. :P

But emotions will be important for a variety of the roles we'll want them to take. For example, one up-and-coming role for robots is one in which they nurse the elderly. I think our elderly will feel much happier being nursed by mecha with emotions and concern for them.

Quote:
Sure i can see the moral issue from an atheistic point of view. I don't believe there is such thing as a soul, so in that sense that makes us kinda equal to higly sophisticated AI. But, in the end of the day an AI will never be anything but 0's and 1's thrown about in different combinations


Sure, they're input/output processors. So are we. Our brain functions on electrical currents as well.

Quote:
unless, of course, scientist one day discover a method on how to create syntethic brain tissue for each an every part of the human brain, but i highly daubt that would ever happens.


No need for this. We can do it with circuits.

Quote:
robots are tools nothing more... Hammers don't have any speciel rights either.


The same has been said of slaves, I'm sure. Humans can be as much tools as machines. And you can't compare something like a hammer, with no intelligence, with complex "artificial" intelligence (I hate the term; intelligence is intelligence).

Quote:
besides... if you could program a highly sophisticated human like AI in a video game... would it be wrong of you to shut down the computer then, or simply quit the game for that matter?


Well, that's just stopping the game. Would it be wrong to take a chainsaw and cut them up? Maybe. I think if we begin putting in ultra complex AI in videogames (even if they lack a material body), there may indeed be something wrong with going "GTA" on them.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Oct 28, 2007 12:03 pm 
Jaden wrote:
Astaroth wrote:
I kinda think it's silly for various reasons.

1: It will never be possible to create an AI equivalent to the human brain, or simulate human behaviour for that matter. Not only is it impossible to program as such for various reasons, but there will never be such a powerful computer either.


You're wrong on both accounts.

First, you don't really give any of these supposed "reasons" why such a thing would never be possible. Humans work on processors too, so I fail to see why there is a limitation on this. I would say more, but you never gave any particular reasons for your statement.


i did later in this thread.

Jaden wrote:
And for your second point, again, I don't know why you would even think such a thing. Especially with the way processor speeds are increasing so rapidly and expontentially.

Humans use massive paralell processing (in the trillions), while machines use one - maybe two in some cases. However, Individual computer processors are way faster than individual human processors. It compensates to some degree. Currently humans are quite a bit more powerful than computers, but there's no reason to think this will always be the case. If there's one technology that isn't seeing a stand-still right now, it's increase in computation speeds.


again... i already covered this. It may be possible with quantum computers but it won't be possible in your lifetime.
And yes, on braincell only has 200 hz, but as long computers can make no more than on or two calculations at a time it will never be able to compensate for that.

Jaden wrote:
If I recall, the processing power of a computer is comparable to that of a mouse (the accumulated sum of its paralell processing). This may not sound impressive, but it is. Consider the evolutionary timescale. Consider the time between the common ancestor of all living beings and the first creature that was as complex as a mouse. It's massive! So, we've come a long way. I mean, our common ancestor would have also been more complex than our first machine too, so you can even expand the analogy further back a bit. As far as processing power is concerned, I would say we're over half way there. Even if we can't replicate trillions of paralell processes any time soon, we can make up for that with faster processors.

I also covered this. And you're wrong. Modern super duber computers are only capable of simulating half a mouse brain run 1/10th of the actually speed. And the difference between a mouse brain and a human brain is quite big.

Jaden wrote:
Quote:
2: there are various forms of AI. AI doesn't necessarily mean they have (real) emotions.


Why would one emotion be real and another fake? What does it matter if the entity is made of metal or meat? Emotion is an object of intelligence, which is an object of processes.

that's not what i meant. An smart AI doesn't need to have emotions.
But they would still be fake if they had emotions. A simulation of reality, nothing more.

Jaden wrote:
Quote:
Why would anybody create a robot with psychical and physical emotions anyway?! Sure, you could have robots say "oh, i love you". I can program my Texas instrument calculator to do that too, yet still.. i would have no trouble throwing in the garbage can one day. You can still program an AI to be creative (i highly daubt it will be anywhere near as creative as a human brain though), and "caring" or "sexy" if that's what needed.


This is just an expansion of what you just said. Again, you don't prove to me that your discrimination is anything but entirely arbitrary.

i'm sorry i'm discriminating your possible future AI :rolleyes:
And you totally didn't get my point.

Jaden wrote:
Quote:
AI will always be a discount version of a real person. Had they real emotions they would commit suicide, rape and burn churches like real human beings. So natural you would have to limit them to be nothing more than smart machines with beneficial functions


If human beings are so bad, then maybe we should stop recreating ourselves. :P

But emotions will be important for a variety of the roles we'll want them to take. For example, one up-and-coming role for robots is one in which they nurse the elderly. I think our elderly will feel much happier being nursed by mecha with emotions and concern for them.


No.. but don't you think prisons for robots would seem rather retarded?!? besides, robots/machinery tends to have supernatural strenght compared to humans, you wouldn't want to mess with them.

Perhaps, but just because you have programmed a "mecha" to show emotions and feel concern for them, doesn't make them real emotions. Like i already stated, i can make my calculator say "i'm concerned about yo. plz take your pills. Oh that's bad Agnes, so you lost your hip the other day.. oh my!" too.
It doesn't make them more real just because your robots have arms, a face and a speaker capable of sounding happy or sad.

besides.. i really don't think elderly care about a robot's concern. A robot may be a good substitute for a wiping elderly ppl in the ass. But real social interaction, i think not.

Jaden wrote:
Quote:
Sure i can see the moral issue from an atheistic point of view. I don't believe there is such thing as a soul, so in that sense that makes us kinda equal to higly sophisticated AI. But, in the end of the day an AI will never be anything but 0's and 1's thrown about in different combinations


Sure, they're input/output processors. So are we. Our brain functions on electrical currents as well.


the difference is, however, that the our brains are under influence of chemicals too.

Jaden wrote:
Quote:
robots are tools nothing more... Hammers don't have any speciel rights either.


The same has been said of slaves, I'm sure. Humans can be as much tools as machines. And you can't compare something like a hammer, with no intelligence, with complex "artificial" intelligence (I hate the term; intelligence is intelligence).

i disagree. Intelligence is not just intelligence, for reasons i already pointed out indirectly in another post. Computer intelligence is still not a valid reason why we should treat them like human beings. It is nothing but software attached to a machine simulating human behaviour.... artificial! artificial! artificial! artificial! Paris Hilton.... artificial... or artificiele in french... kunstig! fake! :cool:

Jaden wrote:
Quote:
besides... if you could program a highly sophisticated human like AI in a video game... would it be wrong of you to shut down the computer then, or simply quit the game for that matter?


Well, that's just stopping the game. Would it be wrong to take a chainsaw and cut them up? Maybe. I think if we begin putting in ultra complex AI in videogames (even if they lack a material body), there may indeed be something wrong with going "GTA" on them.


well.. quiting the game would mean you would cut them from existence. And once you started the game again they would be a different AI since it wouldn't have the same experience as the last one.
Even without a material body you could make them "feel" pain and suffering. That wouldn't make much sense for a material robot to be capable of. The only reason why a Mecha, as you call them, should be able to feel pain would be if you created a Mecha whose intelligence followed the same course as the human behaviour and human identity developes, from baby to adulthood. But since that would complicate things for various reasons, you would have to program an in-build survival mechanism. But in doing so you would also have violated the idea behind the AI's free will and integrity, and it would be nothing but fancy pre-programmed software. How does one program real pain, anyway?


Top
  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 13 posts ] 


Who is online

Users browsing this forum: No registered users and 18 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group